TLDRocket
Sign in

📈 Data to start your week

Exponential View Azeem Azhar

AI data says open models are catching up fast, while young workers in AI-hit jobs are still falling behind. Agents are also eating far more tokens than people now.

Based on reporting by Exponential View, Azeem Azhar — read the original for the full story.

Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error

Exponential View’s Monday roundup points to two forces moving at once: the model mix is shifting toward open weights, and the labour market is still taking hits in places most exposed to AI. Those are not the same story, but they rhyme a little too well.

On the model side, the share of open-weight tokens has doubled over the last twelve months. That puts the system close to a 1:1 split between closed and open tokens. But closed-weight tokens themselves still grew sevenfold in the same period, which is a reminder that “catching up” and “overtaking” are very different things.

Then there’s the labor signal. Employment of young workers aged 22–25 in AI-exposed occupations is now 19% below trend. A year ago, that figure was 15% below trend. The canary, as the note puts it, is still coughing.

The token story is just as stark. AI agents used more tokens than humans in February this year. Since then, agents have expanded to using 14 times that amount, while human token use has grown only 2.8x. That is a very clean picture of where the workload is moving, even before anyone starts arguing about productivity.

Exponential View says it will keep publishing weekly updates on the state of the AI economy based on its own research and tracking, with the latest inference token update feeding into its broader 2026 report.

My take — AI-written commentary, not fact-checked reporting

This is the part of AI everyone keeps trying to dress up as progress: more tokens, more agents, more heat. Open-weight momentum matters, but the labor signal is the one people politely walk around because it’s less fun at conferences. The industry loves automation right up until the bill arrives in someone else’s payroll.

Read more about this at: Exponential View

Related stories

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.